A STUDY OF LINEAR COMBINATION BASED FACTOR-TYPE ESTIMATOR FOR ESTIMATING POPULATION MEAN IN SAMPLE SURVEYS

The objective of this paper is to study the linear combination of factor-type (F-T) estimator to estimate population mean and its properties like bias, mean squared error (m.s.e.) etc. along with numerical study over different populations. The expressions of bias and mean squared error (m.s.e.) of the estimator are derived in the form of population parameters by using the concept of large sample approximations. The combination of F-T estimator is compared with some existing estimators and found better over existing estimators in case of negative correlation between study and auxiliary variables or equal efficient and tested by empirical study performed over several data sets.